Effective Treatment of Respiratory Tract Infections in an Era of Increasing Antibiotic Resistance

Although resistance of Streptococcus pneumoniae to beta-lactam antibiotics was first observed in 1967, it was not until the 1990s that its frequency increased dramatically. During the same decade, resistance of this and other organisms common in respiratory tract infections and otitis media to other antibiotics was also observed, introducing the concept of multi-drug resistance. As alarming as this has been, however, no firm connection has ever been established in respiratory medicine between increased drug resistance and mortality. This apparent absence of a correlation between antibiotic resistance and death rates is due, at least in part, to underemphasizing clinical outcomes when establishing the breakpoints between susceptibility and intermediate
resistance, and between intermediate resistance and extensive resistance, for the Minimum Inhibitory Concentra-tions (MICs) of antibiotics. Recent clinical experience with revising the breakpoints on a disease-specific basis suggests that resistance of
respiratory pathogens to antibiotics, especially the proportion of resistance that results from an intracellular efflux mechanism, may not be as frequent as it has seemed.

These issues of antibiotic resistance and the appropriate use and selection of antimicrobial agents for managing community-acquired bacterial pneumonia were the subjects of an adjunct symposium conducted during the May 2003 conference of the American Academy of Physician Assistants in New Orleans.

This program was supported by an unrestricted educational grant from Abbott Laboratories.

Trends and Issues in Antimicrobial Therapy

Epidemiological studies indicate that Streptococcus pneumoniae, Haemo-philus influenzae, and Moraxella catarrhalis are the dominant organisms in acute otitis media and outpatient respiratory infections including acute maxillary sinusitis, acute exacerbation of chronic bronchitis, and community-acquired pneumonia. S. pneumoniae ranks first in acute otitis media (3% to 35% of cases), acute maxillary sinusitis (25% to 30%), and community-acquired pneumonia (35% to 55%), but is less common than the others in acute exacer- bation of chronic bronchitis (7% to 10%). In light of the prominence of S. pneumoniae in these illnesses, William R. Bishai, MD, PhD of Johns Hopkins University, said that “the sudden appearance and rapid growth of its resistance to antibiotics in the 1990s, after 45 years of reliable susceptibility to antibiotic therapy from the inception of penicillin, was cause for considerable alarm.”

Currently, approximately one third of respiratory pneumococcosis cases are resistant to penicillin and other drugs of the beta-lactam class, and reports reach as high as 60% resistance to penicillin and standard-dose amoxicillin. Clinical studies also indicate that 20% to 25% of cases are also resistant to macrolides and drugs of the tetracycline class, indicating the spread of multi-drug resistance. The pattern of multi-drug re- sistance varies significantly by geographic region. Resistance in varying amounts to amoxicillin is also growing to H. influenzae (24% to 35%) and M. catarrhalis (82% to 97%) depending on the beta-lactamase production by organisms of individual isolates (Thornsberry C et al. Clin Infect Dis 2002;34Suppl 1:S4). Patient response may be improved by the use of a beta-lactamase inhibitor such as clavulanic acid.

The spread of antibiotic resistance is driven by multiple factors. Excessive and/or unnecessary consumption of antibiotics, both clinically and in agriculture, are major causal factors. Long treatment durations, poor compliance, and inadequate dosage regimens are all contributing factors. Treatment failure due to inadequate dosing, in particular, encourages natural selection and reproduction of resistant clones, and promotes their spread in the community. Other factors causing the spread of resistance are the transfer of strains between individuals and transfer of genes encoding resistance between strains. Co-selection of cross-resistant strains also advances resistance.

Accordingly, once a pathogen has been identified, or when empiric therapy is to be initiated, one should first select an appropriate drug class and then use the most potent antibiotic of that class. In addition, because respiratory infections are narrow-spectrum infections, narrow-spectrum antibiotics (macrolides drugs of the tetracycline class) should be used.

Minimum Inhibitory Concentration (MIC) determinations remain the most refined means of measuring in vitro antibacterial activity. MIC is the lowest concentration of an antibiotic that will eradicate a particular pathogen. The lower the MIC, the more potent the drug is against the organism. MIC should not be confused with MIC 50 and MIC 90. MIC is the potency of an antibiotic to a particular strain of a bacterial strain. MIC 50 and MIC 90 represent the minimum concentrations necessary for killing 50% and 90%, respectively, of organisms in a collection of isolates. The National Committee for Clinical Laboratory Standards (NCCLS) uses a variety of broth and agar techniques for determining MICs, MIC 50s, and MIC 90s. It also provides interpretive guidelines for MICs indicating whether an organism at a particular MIC is susceptible (S), intermediately resistant (I), or resistant (R) to the dosage.

Despite the spread of antibiotic resistance among the common respiratory pathogens in the last dozen years, there has been no significant increase in mortality from respiratory infections due to S. pneumoniae, H. influenzae, or M. catarrhalis. Three large studies (N=1,130 to 5,837) covering the periods 1952-62, 1966-95, and 1995-97 reported mortality rates as 13%, 12%, and 12%, respectively (Austrian R, Gold J. Ann Int Med 1964;60:759; Fine MJ et al. JAMA 1996;275:134; Feikin DR et al. Am J Pub Health 2000;90:223). The data came from surveillance studies conducted in academic medical centers and, therefore, involved patients with more complicated respiratory tract infections. These mortality rates bear no relation to the extent of antibiotic resistance. In a study conducted at academic medical centers in 1997-98 involving 1,601 patients in 27 states, 31% of refractory respiratory infections were attributed to penicillin-resistant pathogens and 21% to macrolide-resistant pathogens (Doern GV et al. Emerg Infect Dis 1999;5:757). In a study of 405 patients in 19 states who had community-acquired pneumonia, 56 patients were infected with S. pneumoniae. In that subgroup, the penicillin resistance rate was 23% and that of macrolides was 11%, both substantially below the published epidemiologic data for the same time period (Bishai W et al. Clin Infect Dis 2000;31:229).

The absence of a significant change in mortality during the era of resistance, together with the lack of an established link between in vitro microbiologic resistance and mortality rates and clinical outcomes in patients, constitutes what Dr. Bishai referred to as “the in vitro-in vivo paradox.” Studies conducted during the era of resistance have demonstrated that mortality rates differ very little between penicillin-resistant cases and penicillin-susceptible cases. No study has reported a statistically significant difference in death rates. Possible explanations for this counterintuitive finding are potential alteration in the physiologic state of microbes during infection compared with in vitro conditions, drug activation by the host with potential roles for active metabolites, and immuno-modulatory and anti-inflammatory effects of antibiotics (especially macrolides and fluoroquinolones) that may confer benefits independent of their power to kill microbes (discussed subsequently).

In addition to posing these potential physiologic explanations of the paradox, Dr. Bishai questioned whether or not the MIC breakpoints established by NCCLS, one between S and I and the other between I and R, are accurately set. The criteria used for setting the breakpoints are (i) an analysis of the frequency of MICs and zone sizes, (ii) the presence of known resistance mechanisms, (iii) pharmacodynamic parameters, and (iv) evaluation of clinical outcomes in patients. With regard to pneumococcal respiratory tract infections, the first three criteria have been over-emphasized and clinical outcomes have been under-emphasized. As a result, the Centers for Disease Control and Prevention (CDC) recently asked the NCCLS to raise the breakpoints for pneumococcal respiratory tract infections and otitis media by one dilution but retain the original breakpoints for pneumococcal meningitis. Figure 1 demon- strates the results for ceftriaxone and cefotaxime. The practical implication of this change is that with the new breakpoints, resistance (I plus R) of S. pneumoniae to these agents has dropped from 18% to 4%. To date, the change in breakpoints has been made only for limited representatives of the macrolide class. Thus Dr. Bishai called for “the adoption of disease-specific breakpoints for all of our respiratory tract antibiotics.”

Dr. Bishai illustrated his point using macrolide antibiotics. Approximately 75% of pneumococcal infections in the United States are susceptible to macro-lides. Of the remainder, approximately three-fourths are resistant by way of a concentration-dependent efflux pump which, if overcome by high concentrations of an antibiotic, may become susceptible. The other pneumococcal infections are macrolide-resistant by way of ribosome modification, a process that effectively removes the macrolide’s target and ensures the organism’s non-susceptibility. Macrolides concentrate dramatically in the pulmonary tree, the sinuses, and the middle ear. A blood concentration of 4 mcg/ml yields a lung concentration of approximately 40 mcg/ml, which is sufficient to eradicate intermediately-resistant organisms.

Lonks and colleagues have demonstrated the effect of ribosome modification in a 13-year study of 1,071 cases of pneumococcal pneumonia with breakthrough pneumococcal bacteremia. In their study, 65 patients had strains of S. pneumoniae that were macrolide-resistant, only three of which could be documented as having resistance via the efflux-pump mechanism (Lonks JR et al. Clin Infect Dis 2002;35:556). These findings indicate a very low treatment failure rate with macrolides in efflux- associated resistance in pneumococcal pneumonia. Importantly, whereas the established MIC breakpoints for macro-lides are 0.25 mcg/ml and 0.5 mcg/ml, the clinically relevant breakpoint in this study was 12.0 mcg/ml. In the macrolide class, clarithromycin is the most potent agent, having a five-fold advantage over erythromycin and a four-fold advantage over azithromycin.

Although five expert panels in pulmonary medicine disagree on treatment guideline for respiratory tract infections in the era of antibiotic resistance, one unifying theme is the recommendation of macrolide or doxycycline therapy as first-line empiric management of uncomplicated respiratory tract infections including those with atypical organisms. These agents were tailored for the respiratory tract and are appropriately narrow in spectrum. It is expected that there will be consensus guidelines in 2004. One of the issues the experts will face is the role of fluoroquinolones in empiric treatment of outpatient respiratory tract infections. The very low (0.9% to 1.6%) rate of resistance to fluoroquinolones in the United States has led to a seven-fold increase in the number of prescriptions per 1,000 respiratory tract infections since levofloxacin was introduced in 1997. However, in the winter of 2001-02, pneumococcal resistance to fluoroquinolones more than doubled, suggesting that use may be driving resistance rapidly. It will take more than one season to clarify the trend, of course. Thus far, in comparison trials evaluating fluoroquinolones against clarithromycin, these two antibiotics appear to be equivalent in bacteriologic eradication and clinical outcomes. There are no trial data demonstrating superior efficacy of fluoroquinolones over clarithromycin.

Dr. Bishai suggested that there may be a method other than MIC for explaining the efficacy of macrolides and, to a lesser extent, fluoroquinolones. “Can their immunomodulatory activity explain some of the treatment successes in respiratory tract infections despite microbiologic resistance?” To answer this question, he reviewed evidence regarding the novel immunomodulatory effects of macrolides. They appear to have beneficial effects on neutrophils by increasing accumulation and migration, and by reducing oxidative burst that leads to inflammation. It also appears that they may suppress levels of pro-inflammatory cytokines, have mucoregulatory functions, and improve bacterial clearing through enhanced ciliary activity. Macrolides also appear to inhibit biofilm production, thereby reducing the ability of bacteria to adhere to tissue surfaces.

Excitement has been generated in the treatment of cystic fibrosis following observation in Japan that a patient treated with macrolides experienced objectively improved pulmonary function. In an uncontrolled study of 17 patients that followed, all patients were treated with clarithromycin 500 mg daily for 6 weeks, and experienced mean increases of 14% in forced expiratory volume (FEV1) and 6% in forced vital capacity (FVC). They also had an increased number of sputum neutrophils (Nakanishi N et al. Nihon Kyubo Shikkan Gakkai Zasshi 1995;33: 771; Ordonez CL et al. Am J Resp Crit Care Med 1999;159:A680). In a similar study conducted in the United Kingdom, children treated daily for more than 3 months with azithromycin experienced significant improvements in both FEV1 and FVC (p<0.03 for each) (Jaffe A et al. Lancet 1998;351:420). The improvement was attributed to an immuno-modulatory activity of azithromycin, because macrolides do not have intrinsic antimicrobial activity against Pseudomonas, the dominant pathogen in cystic fibrosis.

Similar benefits have been demonstrated in patients who enrolled in a prospective placebo-controlled open-label study of clarithromycin for chronic sinusitis. Patients were treated with 500 mg of the trial medication twice daily for 14 days, and sinus biopsies taken on days 0 and 21 were compared. Sputum elastic modulus, a measure of viscosity, was measured on day 14, and neutrophil activity and levels of pro-inflammatory IL-8 were assessed on day 8. As Figure 2 indicates, patients had significant improvements in elastic modulus and IL-8 levels (MacLeod CM et al. Adv Ther 2001;18:75).

Melody A. French, PhD, FNP, PA, of the Doyle Family Practice of the Northeastern Rural Health Clinics, Inc. in Doyle, California, addressed issues of antimicrobial use in community-acquired pneumonia from the perspective of someone whose practice is a substantial distance from the nearest hospital, pharmacy, and imaging center.

Based on the recommendations of the American Thoracic Society, the CDC, and the Infectious Disease Society of America (IDSA), Dr. French reiterated that once a diagnosis of uncomplicated community-acquired bacterial pneumonia is made, macrolide or doxycycline therapy is appropriate. Despite the convenience of once-a-day dosing, the fluoroquinolones should be reserved for complicated cases. The choice of a treatment agent is governed by a combination of drug class, potency, dosing convenience and duration, potential adverse effects, and cost. In addition, social factors such as age, literacy, living arrangements including homelessness, social support in the home and the community, and access to reliable transportation should be taken into consideration. These factors are especially important in deciding which patients should be hospitalized. In addition to ascertaining what other medications the patient takes, it is important to inquire about herbal therapies and dietary supplements in order to assess potential adverse interactions with the selected antibiotic.

It is well established that medication compliance in community-based self-medication is directly related to the time spent educating the patient on the nature of her/his illness and the emphasis placed on taking the medication as prescribed and completing the course no matter how quickly normal health appears to have returned. A combination of oral and written instructions maximizes compliance, and written instructions have the added benefit of decreasing the frequency of follow-up phone calls and visits to emergency departments. It is essential that the patient be instructed on what to do if symptoms intensify or fail to abate during treatment and on recognizing adverse reactions (Spiritus E. Am J Manag Care 2000;6[23 Suppl]: S1216).

The successful response of most community-dwelling patients to antibiotics and the general lack of education on the clinical differences between bacterial and viral illnesses lead to frequent situations in which patients with respiratory infections demand antibiotics inappropriately. In these circumstances, it is the clinician’s obligation to make it clear that antibiotics do not benefit viral colds, that the overuse of antibiotics may create serious short-term and long-term health problems including antibiotic resistance, and that viral infections are more effectively treated by symptom management.

Case Presentations
Dr. French presented the case of an 82-year-old male patient with a temperature of 102.2º and a productive cough yielding green sputum. His heart rate was 110 beats per minute and his respiratory rate was 24 breaths per minute at the time of presentation.

He was taking warfarin and a “statin” drug. His atrial fibrillation was stable and his blood pressure and lipids were controlled. He had never smoked, and reported no current alcohol consumption. He had crackling sounds in his left thoracic base and diffuse ronchi. He had a very supportive spouse and a stable home. Because his blood count would not be completed until the next day, and because his wife was unable to transport him for chest x-rays for another day, he was started on empiric therapy with a macrolide. Two days later he appeared for a follow-up visit and was considerably more ill. He was hospitalized and switched to a fluoroquinolone, with rapid success. This case was selected to illustrate the challenges of empiric therapy for respiratory tract infections in patients whose health histories and comorbidities make it unclear if their pneumonia is complicated or uncomplicated.

Dr. Bishai presented the case of a 24-year-old man with a history of childhood asthma and one pack per day of cigarettes for 9 years. He had an elevated temperature, dyspnea, mild chest pain, fatique, and decreased appetite. His leukocyte count was 12,000 and a chest x-ray revealed diffuse interstitial infiltrates in all lung fields. The probable diagnosis for this patient was Mycoplasma pneumoniae, one of the atypical bacterial respiratory tract infections. Beta-lactam agents will not treat this organism, but macrolides, tetracyclines, and fluoroquinolones all will. Dr. Bishai said that a potent member of the macrolide class would be an excellent choice in this instance.